On Sun, 2005-03-20 at 22:22 -0800, Brian Stansberry wrote: > --- robert burrell donkin > <[EMAIL PROTECTED]> wrote: > > inspired by ceki's examples, i've been working on an > > extended systematic > > analysis of the various classloader cases. there's > > demonstration code > > for them as well as documentation. it's not complete > > but most of the > > substantial cases address by ceki's document are > > covered. (the > > unconventional context classloader cases highlighted > > by simon are > > missing but they need a little more consideration.) > > i think that > > developers should find it interesting and maybe a > > little educational. it > > certainly help me think a lot more clearly about > > these issues. > > Look forward to seeing this. I was thinking of > getting going on some unit tests based on the earlier > analyses, but think I'll wait a bit to see what you've > done.
i would recommend waiting. it should be quite easy to add extra analysis cases in the same style to the code base when it's done. i haven't covered much of the ground you had started to explore (nor the ground suggested by simon concerning unconventional classloaders). it would great if these could be added. it may be a day or three before it's ready. i don't particularly want to rush. it's easy to make mistakes when analysing these issues and it's easy to then build misleading arguments upon shaky foundations. i hope that people will bring any mistakes i make to our attention as quickly as possible. i haven't heard any feedback about: http://jakarta.apache.org/commons/logging/tech.html. i'll be relying on the concepts and terminology outlined there in the analysis and demonstration code. again, i'd like to encourage anyone who finds any inaccuracies to speak up as soon as possible. - robert --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
